kde-sig
LOGS
14:02:12 <jreznik> #startmeeting kde-sig -- https://fedoraproject.org/wiki/SIGs/KDE/Meetings/2010-10-19
14:02:12 <zodbot> Meeting started Tue Oct 19 14:02:12 2010 UTC.  The chair is jreznik. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:02:12 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
14:02:25 <jreznik> #meetingname kde-sig
14:02:25 <zodbot> The meeting name has been set to 'kde-sig'
14:02:32 <jreznik> #topic roll call
14:02:39 <jreznik> who's present today?
14:02:59 <Kevin_Kofler> Present.
14:03:23 * rnovacek is here
14:03:58 <jreznik> #chair Kevin_Kofler rdieter than rnovacek
14:03:58 <zodbot> Current chairs: Kevin_Kofler jreznik rdieter rnovacek than
14:04:16 <jreznik> anybody else, than, SMParrish, thomasj...
14:04:23 <rdieter> 1/2 here (1/2 migraine)
14:04:32 * than is present
14:05:42 <jreznik> #info Kevin_Kofler rnovacek jreznik rdieter 1/2 (1/2 migraine), than present
14:06:39 <jreznik> #topic agenda
14:06:52 <Kevin_Kofler> At least: final report for F14, status of KDE 4.5.x and Qt 4.7 for F13.
14:07:43 <jreznik> something else?
14:08:23 <jreznik> nothing, so let's start
14:08:36 <jreznik> #topic final report for F14
14:09:29 <jreznik> 1) Test results for Desktop Validation tests - https://fedoraproject.org/wiki/Test_Results:Fedora_14_Final_TC1_Desktop
14:09:52 <Kevin_Kofler> Well, that's TC1, shouldn't there be RC1 by now?
14:10:37 <Kevin_Kofler> Apparently not yet.
14:10:56 <Kevin_Kofler> From the KDE side, we should be ready though, right?
14:11:10 <jreznik> Test_Results:Current_Desktop_Test
14:11:46 <jreznik> for QA:Testcase_desktop_error_checks I think it was SELinux issue (incorrect reference)
14:11:52 <than> QA:Testcase_desktop_menus still fails
14:12:36 <jreznik> for #643838 - papeli crash - I can't reproduce it
14:13:04 <jreznik> for #643835 - polkit-kde fail - fixed now
14:13:06 <than> jreznik: it seems only produceable in kvm
14:13:49 <than> but it's perhaps a bug in cirrus
14:14:04 <jreznik> but polkit-kde-auth-agent logout crash makes me still worried - I'm not sure it's really fixed now... it's something with gdbus rewrite of polkit
14:14:22 * thomasj late here, sorry
14:14:35 <jreznik> thomasj: yes, probably it is - another qt/cirrus bug
14:14:38 * rdieter can do more f14 testing (esp polkit-kde) after meeting
14:15:05 * ltinkl a bit late as well
14:15:17 <than> jreznik: how can i reproduc the crash in polkit-kde-auth-agent?
14:15:21 <than> jreznik: just logout?
14:15:49 <rdieter> than: do something in your session that uses polkit-kde-auth-agent, then logout, yeah
14:15:50 <jreznik> than: just logout... it happens sometimes, probably some race condition but
14:16:07 <jreznik> rdieter: you even don't need to do any action sometimes...
14:16:26 <rdieter> jreznik: oh, for me I did.  I never got a crash without using it
14:16:36 <rdieter> but maybe I was just lucky
14:17:36 <than> jreznik: i cannot reproduce the crash wit latest f14
14:18:24 <jreznik> it's related to polkit rewrite to gdbus probably... and as we don't know the reasons, how to reproduce, it's not easy to test
14:21:03 <than> it's not easy to fix if it's not reproduceable everytime
14:22:09 * thomasj side note since we're at F14.. Going to update KDevelop to 4.1 final in rawhide and F14
14:22:22 <than> jreznik: can you ask some qa people in brno to test it?
14:23:09 <jreznik> than: lot of people are running plasma f14 here
14:23:58 <than> jreznik: and they cannot reproduce the crash?
14:24:34 <jreznik> than: it happend once for rvokal today and we thought it's fixed by now...
14:26:21 <jreznik> so we're still looking at this issue
14:26:30 <rnovacek> jreznik: i would wait if other people can reproduce it
14:26:51 <rnovacek> it needs logout after update to latest version of polkit-qt
14:27:13 <rnovacek> I don't know if it rvokal's case
14:27:34 <than> we need a working testcase for this.
14:29:12 <jreznik> than: yes
14:29:12 <than> rnovacek: it's great if you and jaroslav just take care of this issue
14:29:58 <jreznik> than: we'd like to... current version just bring too many problems as we're using glib polkit bindings
14:30:22 <jreznik> but I'm not sure we'll have time to rewrite it to clean DBus version
14:30:36 <jreznik> so we have to try to fix current polkit-qt implementation
14:30:50 <jreznik> #info jreznik and rnovacek to work on improving polkit-qt
14:31:04 <jreznik> I'll let you know the status tmrw
14:31:09 <than> jreznik: i think, it's too late for f14, but we can do for f15
14:31:10 <jreznik> anything else for f14?
14:31:41 <jreznik> than: for f14 it's too late but we will have a lot of work on rhel in the near future :(((
14:32:04 <than> jreznik: i know
14:32:24 <than> move on please
14:33:27 <than> i can reproduce the crash in Palapeli
14:33:30 <jreznik> #topic status of KDE 4.5.x and Qt 4.7 for F13
14:33:32 <than> in kvm
14:33:49 <jreznik> than: in kvm? with cirrus? try vmware vga video pls
14:33:51 <than> it's cqt/irrus issue again
14:34:20 <than> jreznik: i don't think it will crash with other driver
14:35:42 <rdieter> I guess we can get started doing kde-4.5.x for f13.  still waiting for approval for a f12 buildroot to do an unofficial update there too
14:35:53 * than is trying vmware vga
14:35:54 <rdieter> as for qt-4.7.x, not sure.
14:36:32 <thomasj> 4.5.2 and Qt-4.7 together seems to work very well.. At least here in my F13 box
14:36:33 <than> rdieter: do we already have dist-f13-kde target?
14:36:42 <rdieter> (fwiw, rumors are that 4.7.1 is right around the corner)
14:37:01 <rdieter> than: not yet, but I can enable it easy enough
14:37:17 <thomasj> Just, lets take care of Yawp
14:37:45 <than> rdieter: we should enable it asap if someone of us wants to build
14:37:46 <rdieter> than: added now, the initial repo generation takes a fair amount of time
14:38:03 <than> rdieter: cool
14:38:09 <ltinkl> yup, can only confirm that 4.5.2 and Qt-4.7 work nicely together
14:38:57 <than> yes, 4.5.2 and 4.7 works fine here too
14:39:15 <jreznik> worked for me too but I'm now on F14
14:39:26 <rdieter> I would personally lean toward only doing kde-4.5.x, and omit qt-4.7.x for now, but that's just me.  adding in qt-4.7.x will just mean a bit more work, and rebuilds (like PyQt4 I think)
14:39:56 <rdieter> that's what we've been testing in kde-testing repos so far (building against only qt-4.6)
14:40:27 <rdieter> but... we do already have qt/f13 branch up'd to qt-4.7
14:40:42 <rdieter> (not that it would be a big deal to revert it, but...)
14:40:56 <rdieter> so... sounds like preferences are to to qt-4.7 too?
14:41:02 <rdieter> to do...
14:41:12 <thomasj> yes
14:41:30 <rdieter> (though that goes a bit against our update proposal that was brought before fesco awhile back)
14:41:47 <rdieter> where several fesco members were ok with a kde update, provided qt was not included
14:41:50 <than> rdieter: yes, we will do qt-4.7 upfate for f13
14:42:10 <thomasj> Do we have any Qt related apps that might get hurt with a Qt update to 4.7?
14:42:17 <thomasj> Erm
14:42:27 <thomasj> Qt apps outside of the KDE SIG
14:42:44 <than> thomasj: it's compatiable to 4.x
14:42:54 <rdieter> thomasj: not aware of any off the top of my head, but their being outside our purvue, means we're not 100% sure either
14:43:00 <than> i don't think it will cause problem
14:43:03 <thomasj> So should be no problem to update to 4.7 then
14:43:15 <jreznik> rdieter: the new rules apply for f14+, don't they?
14:43:37 <rdieter> jreznik: good question... let's go with 'yes'. :)
14:44:05 <jreznik> but yes - qt update can probably broke more than clean kde update...
14:44:18 <jreznik> even it should be compatible but you know...
14:44:55 <jreznik> it would need headups for all qt maintainers
14:45:22 <thomasj> True.
14:45:41 <jreznik> on the other hand - KDE SIG is responsible for Qt, so we should be masters here and say yes/no
14:46:03 <rdieter> ok, several steps involved here:  1.  do a devel-announce post to publicize our intentions
14:46:05 <than> yes, every update can have regressions ....
14:46:18 <than> it needs to be tested well
14:46:29 <jreznik> it could help a lot for example arora and rekonq
14:46:31 <than> before we push it into update
14:46:40 * thomasj tests 4.5.2 with QT 4.7 all the time it's out
14:46:45 <jreznik> as we would have newer qtwebkit
14:46:49 <rdieter> 2. do builds, test test test
14:47:07 <jreznik> rdieter: agree and in case of objections we will just update kde only
14:47:26 <rdieter> jreznik: too bad qtwebkit isn't separate (yet).  someday maybe.
14:48:06 <rdieter> (ah... joy... my migraine medicine is starting to work wonders)
14:48:19 <thomasj> Or we satisfy our users of F13 and give them finally KDE 4.5.2 as update. They wait since a long time.
14:48:32 <thomasj> rdieter: woo, medicine ftw :)
14:48:58 <thomasj> And leave Qt-4.7 out.
14:49:08 * thomasj scratches head and dislike FESCo once more
14:49:13 <jreznik> thomasj: we can be much more faster without Qt, you're right
14:49:49 <rdieter> :)  are we waffling on qt47 now?
14:50:01 <thomasj> yup :)
14:50:18 * jreznik is checking "waffling" in dictionary
14:50:44 <rdieter> jreznik: changing your mind (and often going back-n-forth)
14:51:01 <rdieter> that's my rough intention for using the term anyway
14:51:47 <than> we should consider the security issue in qtwebkit
14:51:50 * ltinkl like waffles
14:51:57 <thomasj> :D
14:51:57 <rdieter> ltinkl: +1, mmm
14:51:59 <ltinkl> likes even
14:52:04 <than> qt-4.7 has new qtwebkit
14:52:04 <jreznik> wow, waffles
14:52:12 <rdieter> than: oh, there is?  that may change things then
14:52:19 <jreznik> than: yes, it's a big advantage of doing qt 4.7 update
14:52:27 <than> it's easy to backport fixes
14:52:33 <rdieter> oh, ok.  true.
14:52:36 <than> if we have qt-4.7
14:53:04 <rdieter> ok, consider me officially waffling closer to wanting to do qt47 too then. :)
14:53:10 <than> jreznik: yes, it's why i want  qt.4.7
14:53:14 <thomasj> We tested 4.5.2 with Qt-4.7 already, right? So we don't have to do weeks of testing again, right?
14:53:28 <jreznik> I think we should go the way rdieter suggested - let others know our intentions to update qt - in case of no objections - I would say - go with new Qt 4.7 for f13
14:53:31 <rdieter> thomasj: yes and no.  for f14 yes.  for f12/f13 builds, not as much
14:53:35 <thomasj> Just a normal get it built and into updates-testing for the masses
14:54:04 <ltinkl> exactly, I've had 4.5.2 with Qt 4.7 already for some time now, with no problems
14:54:10 <rdieter> ok, so... who wants to make the announcement (or lead the effort to draft one anyway)?
14:54:16 <jreznik> in f14 all qt updates are probably bad idea as the new evil update policy comes to eat us and still better something (kde) than nothing
14:54:17 <thomasj> I run 4.5.2 with Qt-4.7 since a while, had even 4.5.1 with it
14:54:53 <rdieter> jreznik: yeah, but we're at a point where qt/kde is stable and featureful enough that updates are less urgent too
14:55:08 <thomasj> At least two hardcore testers, ltinkland me ;p
14:55:09 <rdieter> we'll se
14:55:20 <thomasj> *ltinkl and me
14:56:15 <rdieter> I can do the announcement, if needed, but would prefer is someone else could do it this time (it's good experience and practice for collaborating)
14:57:03 * ltinkl is still busy with HALsectomy
14:57:08 <jreznik> rdieter: and it's sad that new KDE updates are not getting more conservative but actually opposit ones...
14:57:30 <jreznik> rdieter: I'll do it
14:59:06 <jreznik> #info jreznik to do a devel-announce post to publicize our intentions regarding kde/qt update in f13
14:59:16 <jreznik> we are nearly out of time
14:59:50 <jreznik> so let's see what will happen :) and if we will see any objections from other qt apps maintainers
14:59:54 <rdieter> guess so, I think we covered all agenda topics
15:00:26 <jreznik> back to the polkit-kde lougout crasher - it was blocker and we're not sure completely if it is really fixed...
15:00:45 <jreznik> let's move to #fedora-kde
15:01:00 <jreznik> thanks for meeting
15:01:01 * ltinkl is happy to have avoided using udisks/upower thru the glib bindings
15:01:28 <jreznik> ltinkl: you're going through DBus?
15:01:32 <jreznik> #endmeeting